alex

agree with Kerr

I tend to agree with Kerr on this one. (Not that that matters much because I am at present (and perhaps forever) an anonymous nobody). Nevertheless -- I think one of the flaws in the above discussion is the idea that one type of exam fits all types of courses. A few examples:
In my constitutional litigation class (which is a "quasi fed courts class," taught at UPenn -- heavy emphasis on Section 1983, immunities, 11th Amendment, Ex Parte Young, Abstention, etc. but without any discussion of habeas) - the exam was a 24 hour take home exam. I thought that this type of exam perfectly suited the course (which in the end consisted of somewhere around 3000+ pages of unedited Supreme Court cases) because it was the own fair way to even touch upon a significant fraction of the material, . The exam felt like this: you have one day to get back to a sr. associate/partner an answer to two fact patterns that raise a host of constitutional litigation/fed courts issues (I think the test had 2 fact patterns, each consisting of 4 sub-parts). And apropos for this course, there was also third "policy" question. The exam for sure took the full 24 hours to write, re-read, edit, etc. My answer, if I recall, turned out to be around 25-30 pages. So it was a lot of work for one day. But as I mentioned, I thought this was fair because it gave students an opportunity to demonstrate mastery across some (or all) of the course, whereas an in-class exam (whether 3, 4, or 6 hrs) would have undoubtedly only touched upon a narrow portion of the class. The professor was also able to digest these long answers because the class was highly self-selective and very small (only 15 or so students).


Now, on the other hand, I thought using a hybrid of multiple choice questions and short essays for classes like evidence and constitutional criminal procedure were ideal. Both of these subject matters are almost purely "rule-orientated," and the point - or at least I think so - is to test whether you know lots of discrete rules and are able to apply them correctly in situations that are likely to repeat themselves over and over and over (i.e. hearsay objections, miranda, etc.). And so they lend themselves well to being tested vis-a-vis multiple choice questions. (Multiple choice also works well because these aren’t “doctrinal” courses per se – here I think there’s less of a focus in testing whether the student understands how the law “evolved. But I don’t want the takeaway being that these classes could not be taught in a theoretical manner, but my experience is that, at least for the introductory versions, they aren’t - which I think is appropriate because these classes severe to have a much more “practical import” than say the first year survey courses, with the exception of civil procedure). That said, the use of some short essays worked well because it simulated a motion in limine (and frankly made the test a little more fair to those who don't necessarily excel at multiple choice exams).